Jakarta, INTI - SK Hynix Inc. reportedly plans to double its memory chip production capacity over the next five years. This major expansion is expected to help alleviate the global shortage of memory chips, a crucial component in the development of artificial intelligence (AI).
The South Korean company is responding to a storage chip supply shortage that is expected to persist until 2030. SK Hynix Chairman Chey Tae-won told reporters in Taipei that the company is maintaining its previous projections regarding the supply deficit.
SK Hynix is increasing its investment spending to address the imbalance between demand and supply. However, Chey continued, the exact investment amount is difficult to determine because the prices of various resources such as land, equipment, and electricity continue to fluctuate. He emphasized that SK Hynix will provide whatever funding is needed to expand its chip wafer production capacity.
"Whatever we need, we will provide. There will still be a supply shortage until 2030," Chey said.
SK Hynix, along with Samsung Electronics and Micron Technology, dominates the global memory market and is one of the biggest beneficiaries of the global AI data center boom. Trillion-dollar spending projections by hyperscalers like Meta Platforms pushed SK Hynix and Micron's market valuations past US$1 trillion for the first time last week.
Memory chips, used to store and process the vast amounts of data required for AI services, are now one of the biggest bottlenecks in AI development. Investors and analysts expect the memory shortage to persist until at least 2027, giving memory manufacturers greater pricing power over the world's largest technology companies.
Some industry executives have even called the current situation a "supercycle," driven by surging demand for high-bandwidth memory (HBM), used to train AI models like ChatGPT, as well as more conventional DRAM and NAND Flash used in data centers.
This phenomenon has helped the memory industry emerge from a decades-long boom-and-bust cycle and boosted the South Korean economy.
COVID-19 Pandemic Slowed Down Expansion
On Tuesday, Arm Holdings CEO Rene Haas stated that memory remains the biggest bottleneck for the global AI industry. He explained that one reason is that industry leaders temporarily reduced capacity expansion during the post-COVID-19 pandemic slowdown several years ago.
However, the memory industry remains cautious about over-expansion, which could trigger a repeat of the previous market crash.
Chey added that building new production facilities is quite time-consuming. Greenfield projects can even take more than five years to become operational. Previously, SK Hynix executives stated that the company's capital expenditure in 2026 would increase significantly compared to the KRW 30.2 trillion budgeted for 2025, although they did not provide further details.
Furthermore, SK Hynix has also filed for a stock listing in the form of American Depositary Receipts in the United States this year. If implemented, the move could potentially be one of the largest debuts of a foreign company on the New York Stock Exchange and give American investors greater access to invest in the AI memory chip business.
Conclusion
SK Hynix plans to double its memory chip production capacity in the next five years to overcome a global supply shortage that is expected to last until 2030. Along with Samsung Electronics and Micron Technology, SK Hynix is one of the main beneficiaries of the surge in global AI data center construction. High demand for high-bandwidth memory (HBM), DRAM, and NAND Flash for AI needs is pushing the memory industry into a strong growth phase or “supercycle”. However, the company remains careful not to over-expand that could trigger a market crash like in the previous cycle.
Read more: Samsung to Build Chip Testing Plant in Vietnam, Slated to Start Operations in November 2027